Jin Dynasty population.

Sep 2012
1,121
Taiwan
Well it's generally agreed the total population in 1100 (N. Song) and 1200 (S. Song + Jin) was roughly the same; around 120 million. So population overall was relatively static, and there wasn't a boom as there had been in the 11th century, which is why I've been a little sceptical about any sudden or major growth (or even decline) on either side of the Yangtze here.

To be honest, the more I read, the more I'm convinced there is some issue with Kuhn's numbers and/or reasoning. If you can read Chinese, the ZGRKS is good place to start, although Durand's The Population Statistics of China, A.D. 2-1953 and Deng's Unveiling China's True Population Statistics for the Pre-Modern Era with Official Census Data are both worth looking at too (in English). Deng actually does show a supposed rise in population based on census figures in this period (which is what you've presumably stumbled upon*), but demonstrates that, using adjusted figures, population overall actually slightly decreased in the twelfth century (p.45, fig.6). He doesn't list the individual dynastic differences (taking instead a combined figure), but he still demonstrates (reasonably persuasively) that there really wasn't a massive loss of life during the Jin conquest, nor any subsequent boom to make up for it. Granted people assuredly did lose their lives, but talking about tens of millions puts us in Mongol and Manchu conquest territory, which were certainly on a far larger scale.



*Edit: on reflection, it appears he's actually taking the male-only Song census data, so the rise here can actually be explained by the addition of the full Jin census data.
 
Last edited:
Sep 2012
1,121
Taiwan
Okay, crunching a few more numbers and running over some more sources:

North/south population split in ZGRKS:

Population above Huai River (i.e. (future) Jin territory):
1102: 35.5 million (~36%)
1200: 44 million (35%)

Population below the Huai River (i.e. S. Song territory):
1102: 63 million (64%)
1200: 80 million (~65%)

Proposed total populations:

1102: ~99 million
1200: ~124 million

//

Deng's adjusted population figures, total population:

1100: 115 million
1110: 120 million
1187: 110 million
1190: 111 million
1195: 112 million

//

Population from census data:

Jin:
1161: 19.5 million (*not census, extrapolated from Jinshi) / 33 million (ZGRKS, proposed estimate)
1187: 44 million
1190: 45 million
1195: 48.5 million
1207: 53.5 million

Song (rough estimates by moi, just doubling male population*):
1100: 90 million
1102: 90 million
1110: 92 million
1161: 48 million
1162: 66 million?
1178: 57 million
1180: 54 million
1187: //Kuhn estimates 61.8 million
1193: 56 million (Durand gives rough estimate in this period for 75 million)
1223: 56 million

*Can't find anyone currently to have given a comprehensive list of adapted figures, per census. Extrapolating from households would be more accurate, but I don't like math. Would seem these figures are maybe off by 10-20 million, given more professional estimates by others.

So a number of points:

1) I will hold my hands up and apologise; people other than Kuhn have noted a population drop in the early S. Song, so I was perhaps a little too quick to write that off! However...

2) Kuhn was likely underestimating when he said 25% in north China prior to Jin conquest, given data in the ZGRKS. I've finally managed to look at the book and it's interesting to see that he doesn't cite much in the way of population statistics or theory to support his arguments either though. However, he also doesn't out and out trust these numbers either, and does advise to treat them with caution; he's certainly not chosen this hill to die on, as it were.

3) There are a lot of numbers! And a lot of ways of interpreting them and extrapolating from them, some more accurate than others (i.e. not my own). There are, as such, no 'best' or most authoritative numbers (although the Jin figures are largely agreed to be accurate, post 1187), and no one really agrees on everything exactly.

4) As above, there could indeed be an issue regarding a 'missing' part of the S. Song population after the Jin conquest. If we take Kuhn's S.Song estimate for 1187 (61.8 million), as he suggests in his book we get to around 105 million total by adding in the Jin population (44 million), which is just shy of Deng's adjusted total (110m) based on some more serious maths. So this seems like the best yardstick. The problem, however, is that Kuhn is likely still underestimating the original N. Song population above the Huai River. He puts this at 25 million, when actually it may well have been closer to 35 million (which is still potentially conservative, as the ZGRKS total population estimate is largely below others). When we adjust to this estimate, the numbers correlate much better, and we don't have a shortfall to make up (1100 N: 35m > 1187 N: 44m // 1100 S: 63m > 1187 S: 62m). As such, there was probably no population 'boom' in the Jin, as they inherited a much larger population than Kuhn believes (which again doesn't seem to factor in Liao and Jurchen populations).

5) However, we are still faced with the peculiar loss of S. Song people, as their population was largely halved from 100 million (for simplicity's sake) to around 50 million in the mid twelfth century, whereas the Jin population was below this, with a safe estimate of around 30 million people.* So whilst the population has evened out by 1187 and returned to 'normal', we are missing around 20 or so million people in the period c.1100 - c.1160. As mentioned above, it has been a problem examined by a few historians, who have likewise pointed out some of the issues and questions raised here. So where did they go? Naturally we might consider that some lost their lives in the invasion and others fell off the grid as Song census standards declined. However, migration also ostensibly played a big part. Whilst 20 million is certainly on the high side, some kind of net swing of migrants from the Song to the Jin in the mid twelfth century seems likely, which helps further explain the sharp population loss in the S. Song and the sudden population gain in the Jin. So actually Unknown, you may have been right in your earlier idea, apologies for not giving credence to it sooner! Deng Gang in his The Premodern Chinese Economy goes into some detail about this shift in internal migration (pp.311-314), attributing it mainly to supposed early unpopular S. Song tax policies instituted to prop up the new regime, the loss of political and social authority, as well as favourable agricultural and tax situations in the north.

*Naturally people disagree on the amount; one scholar even states that the 1187 population was triple that of the initial Jin population. Whether he meant the 1115 population or the 1127 population I do not know, although if it were the latter, it would certainly be one of the lowest estimates. Given data and proposals in the ZGRKS, as well as factoring in that the Jinshi extrapolation of 19.5m is probably on the conservative side, 30m seems likely to me.

Apologies for retreading over a lot of stuff and backpeddling a bit on others; I'm learning about this on the fly, so I'm still trying to figure it all out myself!
 
  • Like
Reactions: Ichon
Sep 2018
46
Germany
Okay, crunching a few more numbers and running over some more sources:

North/south population split in ZGRKS:

Population above Huai River (i.e. (future) Jin territory):
1102: 35.5 million (~36%)
1200: 44 million (35%)

Population below the Huai River (i.e. S. Song territory):
1102: 63 million (64%)
1200: 80 million (~65%)

Proposed total populations:

1102: ~99 million
1200: ~124 million

//

Deng's adjusted population figures, total population:

1100: 115 million
1110: 120 million
1187: 110 million
1190: 111 million
1195: 112 million

//

Population from census data:

Jin:
1161: 19.5 million (*not census, extrapolated from Jinshi) / 33 million (ZGRKS, proposed estimate)
1187: 44 million
1190: 45 million
1195: 48.5 million
1207: 53.5 million

Song (rough estimates by moi, just doubling male population*):
1100: 90 million
1102: 90 million
1110: 92 million
1161: 48 million
1162: 66 million?
1178: 57 million
1180: 54 million
1187: //Kuhn estimates 61.8 million
1193: 56 million (Durand gives rough estimate in this period for 75 million)
1223: 56 million

*Can't find anyone currently to have given a comprehensive list of adapted figures, per census. Extrapolating from households would be more accurate, but I don't like math. Would seem these figures are maybe off by 10-20 million, given more professional estimates by others.

So a number of points:

1) I will hold my hands up and apologise; people other than Kuhn have noted a population drop in the early S. Song, so I was perhaps a little too quick to write that off! However...

2) Kuhn was likely underestimating when he said 25% in north China prior to Jin conquest, given data in the ZGRKS. I've finally managed to look at the book and it's interesting to see that he doesn't cite much in the way of population statistics or theory to support his arguments either though. However, he also doesn't out and out trust these numbers either, and does advise to treat them with caution; he's certainly not chosen this hill to die on, as it were.

3) There are a lot of numbers! And a lot of ways of interpreting them and extrapolating from them, some more accurate than others (i.e. not my own). There are, as such, no 'best' or most authoritative numbers (although the Jin figures are largely agreed to be accurate, post 1187), and no one really agrees on everything exactly.

4) As above, there could indeed be an issue regarding a 'missing' part of the S. Song population after the Jin conquest. If we take Kuhn's S.Song estimate for 1187 (61.8 million), as he suggests in his book we get to around 105 million total by adding in the Jin population (44 million), which is just shy of Deng's adjusted total (110m) based on some more serious maths. So this seems like the best yardstick. The problem, however, is that Kuhn is likely still underestimating the original N. Song population above the Huai River. He puts this at 25 million, when actually it may well have been closer to 35 million (which is still potentially conservative, as the ZGRKS total population estimate is largely below others). When we adjust to this estimate, the numbers correlate much better, and we don't have a shortfall to make up (1100 N: 35m > 1187 N: 44m // 1100 S: 63m > 1187 S: 62m). As such, there was probably no population 'boom' in the Jin, as they inherited a much larger population than Kuhn believes (which again doesn't seem to factor in Liao and Jurchen populations).

5) However, we are still faced with the peculiar loss of S. Song people, as their population was largely halved from 100 million (for simplicity's sake) to around 50 million in the mid twelfth century, whereas the Jin population was below this, with a safe estimate of around 30 million people.* So whilst the population has evened out by 1187 and returned to 'normal', we are missing around 20 or so million people in the period c.1100 - c.1160. As mentioned above, it has been a problem examined by a few historians, who have likewise pointed out some of the issues and questions raised here. So where did they go? Naturally we might consider that some lost their lives in the invasion and others fell off the grid as Song census standards declined. However, migration also ostensibly played a big part. Whilst 20 million is certainly on the high side, some kind of net swing of migrants from the Song to the Jin in the mid twelfth century seems likely, which helps further explain the sharp population loss in the S. Song and the sudden population gain in the Jin. So actually Unknown, you may have been right in your earlier idea, apologies for not giving credence to it sooner! Deng Gang in his The Premodern Chinese Economy goes into some detail about this shift in internal migration (pp.311-314), attributing it mainly to supposed early unpopular S. Song tax policies instituted to prop up the new regime, the loss of political and social authority, as well as favourable agricultural and tax situations in the north.

*Naturally people disagree on the amount; one scholar even states that the 1187 population was triple that of the initial Jin population. Whether he meant the 1115 population or the 1127 population I do not know, although if it were the latter, it would certainly be one of the lowest estimates. Given data and proposals in the ZGRKS, as well as factoring in that the Jinshi extrapolation of 19.5m is probably on the conservative side, 30m seems likely to me.

Apologies for retreading over a lot of stuff and backpeddling a bit on others; I'm learning about this on the fly, so I'm still trying to figure it all out myself!
Thank you for the elaborate answer!


A mass migration from south to north would contradict most things we do know about thes outhern Song, as well as about the overall pattern in Chinese history of a population shift south. The Jin regime must have had vastly superior living standarts to have changed that pattern. While I'm sure that some of the refugees from the war did return north after the Jin restored order there, I cant believe that it was so many. Also Cambridge 5 does write about agricultural problems in the Jin and their dependency on Rice imports from the south. The fact that the north was war torn but the population did decline in the south (and its decline is noted in the early Southern Song census, while it only grew in the following censuses) is strange, since if there was any migration to theJin, it would have taken place after the Jin got relatively sinicised and can't explain the early drop. As for the population of China recovering to pre Sogn-Jin war levels only by 1200, this could only be true, if there was a drop with the Jin conquest in the first place. If there was none then the population must have continued to grow. Also there are some scholars who do argue for a continuing population expansion in the Southern Song period. Kuhn does make the suggestion, that the population by 1200 might have been far larger, due to great undercounting (probably in the southern Song) and that half the worlds population lived in China by1200. There is also Banister, whos figures are in wiki: Demographics of China - Wikipedia
and who speaks of 140 million by 1200, then again she also believes in a Yuan population explosion and 120 million people by 1351 (with the black death having been ravaging China for over a decade by that point), so Id take her numbers with a grain of salt. Still both Jin and Song censuses do show natural population growth after 1180,so I see no reason not to believe that Chinas population by 1200 was larger then in 1100,if there was no massive depopulation in the Jin conquest.
 
Sep 2012
1,121
Taiwan
Hmm, Song migration to the Jin is very likely to have occurred in the early twelfth century; the S. Song was in turmoil and had just lost a hell of a lot of legitimacy, lacking both the clout and means to keep people in the south. Furthermore, it hit most heavily southerners who had largely managed to avoid excessive taxation up until then; the relocation of the court to the south and the new tax levies raised to support the administration and the army would have certainly forced people north. The Jurchens offered lower taxes, lower prices and other amenities which would have made migration incredibly attractive. I'm not sure what sinicisation has to do with anything; they were moving to Chinese cities with Chinese people. Concepts of loyalism in this period are usually (although of course not always) more to do with loyalty to the state, rather than loyalty to an ethnocultural group; your average person on the street probably didn't care what ethnicity the emperor was, provided taxes were low and crop yields were high. So yes, northward migration would buck the historical trend, but it is an understandable anomaly.

It is clear ... that from the early eighth century to the fourteenth century, the overall direction of internal migration in the long term was to the south. What seems puzzling is that the tide changed in the opposite direction during the Southern Song Period when the Tartar Jin and Southern Song régimes coexisted. First, considering the coexistence of these two régimes, one would think that it would have been natural for the Han Chinese population to have packed into the south to avoid the alien conquerors in the north. Second, considering that the south was not a war zone, one would expect that the south not to have lost a large proportion of its population. Third, considering the violent process of alien conquest, one would imagine that north China would lose a sizeable proportion of its population and that its recovery would have needed time. Yet by 1187 the northern population under the Tartars had increased over threefold from its pre-conquest level in 1102 with a minimum annual increase rate of 1.3 per cent, as if the war had never taken place. If the population loss during the war of conquest is taken into account, the rate of increase in the north is certainly much greater. Comparatively, in the south, where peace was to a great extent retained, the population dropped by 46 per cent in 1102– 59, with an annual negative growth rate of 1.4 per cent. Percentage wise, the gain and loss matched nicely, maybe too nicely. But if the natural population growth rates in the north and south were similar, and thus constant, this would happen. The dramatic population shift thus indicates two things: first, the loss of the Southern Song population was to a great extent a gain of the Tartar Jin; second, the reverse migration of the Chinese population to the north was voluntary and ethnicity was not a consideration.

At least four key factors, two from the Song side and two from the Tartar side, contributed to the reverse population shift. First, owing to the structural changes during the Northern Song Period, China was weakened and consequently lost its northern half. The Song’s humiliating failure to defend China along the Great Walls dramatically changed the image of the indigenous government in the eyes of the Chinese population, which seriously undermined the government’s credibility, legitimacy and authority. What made it worse was that in 1129, to escape from the Tartars, Emperor Gaozong (r. 1127– 62) exiled himself with a fleet from Mingzhou and remained at sea for four long months ...The Song monarch in effect abandoned the empire in order to save his own neck. This was the ultimate betrayal of his Chinese subjects and a total violation of the Confucian rule of ‘People as the Foundation’ (minben). This was particularly serious given the overcentralisation of the Song state, with which the emperor’s responsibility increased exponentially. It was a miracle that the Southern Song did not collapse within these crucial months.

Second, when the exiled Song government withdrew to the south, a large number of bureaucrats and troops went with it. The old problems of overspending and budget deficit recurred under the Southern Song. Evidence shows that by 1196 the Southern Song government payroll had some 42,000 officials and over 400,000 military personnel. The total wage bills for theses two groups reached 5,000 million and 80,000 million bronze coins, respectively, considerably higher than under the Northern Song. In addition, to resist further Tartar aggression, the Southern Song urgently built a new defence line along the Yangzi River... Conceivably, all these expenses were translated into a tax burden on the residents in the south...

With the southward move of the state apparatus, Song control over the south intensified, which brought the southern tax-evasion haven to an end. Meanwhile, the south was infected by the landholding concentration disease ... This was not all: the Southern Song government joined in and scrambled for land. By law, the state had the power to purchase up to one-third of the estates that fell into certain categories. Under this scheme, the state brought a total of 203,000 heactares from six prefectures to yield rents. The payments were made by the Song government with paper currency. Given the way the Song paper currency was issued, this was close to land robbery...

Third ... the war of the Tartars against the Song in North China eliminated the entrenched Song administration and influential officials and large landowners, vested interest groups who proved to be less tolerable to the Chinese than the alien conquerors. The adoption of the trinary structure in the Tartar-controlled north, marked by the ‘Land Allocation According to Clan Members’, made remigration back to the north more attractive.

Last, the Tartar Jin made the Chinese peasantry feel relatively comfortable: the regime actively encouraged farming through (1) the meng-an moke system, characterised by Tartar land ownership and Chinese tenancy; (2) a reasonably fair tax system copied from the Tang output-based ‘Two Seasonal Taxes’ (liangshui fa); and (3) systematic confinement of merchant activities, also copied from the Tang.

Consequently, the direction of internal migration began to reverse, given the unprecedented distrust of the Song government, and for the same purposes of tax evasion and landholding. So, the unthinkable happened: large numbers of Chinese defected from the Southern Song regime and joined the Tartars. The fact that many of the immigrants were tenants and traders made their reverse migration easier since the opportunity costs for the landless or nearlandless to move were normally lower than those for landowners. That so many people managed to emigrate back to the north indicates that the Song government was further weakened. This was reflected by the quality of the Southern Song census, which was considerably poorer than that of the Tartar Jin.
Deng, Gang (1999), as above, pp.311-314. Sorry, I tried to abridge it as best I could, but this is his reasoning for northward migration.

As for loss of life during the Jin invasion and a subsequent mini-boom in population back to pre-war levels, I'm still a little sceptical. Deng here certainly accepts that theory to a certain degree, but he also gives a much lower estimate of early Jin population than others. I'm more inclined to think that whilst there was assuredly some loss of life, it can't have been on a comparable scale to other invasions or conflicts which we know historically did far more damage, so a fall and rise in census figures to my mind more likely obscures poor record keeping, mass displacement and fluctuating migration across the border. I could of course be wrong, but that's my working theory, based upon what I've observed so far; certainly open to convincing though!

I'd be interested to know where Banister gets 140m from for 1200, that's certainly the highest figure I've seen! As for a larger population in 1200 than 1100, I haven't seen any evidence for that in any study, bar the isolated example given above in the ZGRKS for north/south estimates (which I don't entirely understand, given that I'm pretty sure it gives a higher 1100 total population elsewhere in the book). Most population scholars seem fairly sure the overall population remained largely the same, give or take perhaps 10 million. I'm most inclined to believe Deng's adjusted figures (given above), as they seem to be the most thorough and accurate; roughly 115 million for both dates.
 
Sep 2018
46
Germany
Hmm, Song migration to the Jin is very likely to have occurred in the early twelfth century; the S. Song was in turmoil and had just lost a hell of a lot of legitimacy, lacking both the clout and means to keep people in the south. Furthermore, it hit most heavily southerners who had largely managed to avoid excessive taxation up until then; the relocation of the court to the south and the new tax levies raised to support the administration and the army would have certainly forced people north. The Jurchens offered lower taxes, lower prices and other amenities which would have made migration incredibly attractive. I'm not sure what sinicisation has to do with anything; they were moving to Chinese cities with Chinese people. Concepts of loyalism in this period are usually (although of course not always) more to do with loyalty to the state, rather than loyalty to an ethnocultural group; your average person on the street probably didn't care what ethnicity the emperor was, provided taxes were low and crop yields were high. So yes, northward migration would buck the historical trend, but it is an understandable anomaly.



Deng, Gang (1999), as above, pp.311-314. Sorry, I tried to abridge it as best I could, but this is his reasoning for northward migration.

As for loss of life during the Jin invasion and a subsequent mini-boom in population back to pre-war levels, I'm still a little sceptical. Deng here certainly accepts that theory to a certain degree, but he also gives a much lower estimate of early Jin population than others. I'm more inclined to think that whilst there was assuredly some loss of life, it can't have been on a comparable scale to other invasions or conflicts which we know historically did far more damage, so a fall and rise in census figures to my mind more likely obscures poor record keeping, mass displacement and fluctuating migration across the border. I could of course be wrong, but that's my working theory, based upon what I've observed so far; certainly open to convincing though!

I'd be interested to know where Banister gets 140m from for 1200, that's certainly the highest figure I've seen! As for a larger population in 1200 than 1100, I haven't seen any evidence for that in any study, bar the isolated example given above in the ZGRKS for north/south estimates (which I don't entirely understand, given that I'm pretty sure it gives a higher 1100 total population elsewhere in the book). Most population scholars seem fairly sure the overall population remained largely the same, give or take perhaps 10 million. I'm most inclined to believe Deng's adjusted figures (given above), as they seem to be the most thorough and accurate; roughly 115 million for both dates.
Ok I did check the Cambridge history of China for both Jin and Song and they dont mention any migration north, instead they write this:

"In the wake of the Jurchen’s brutal destruction of human life and urban wealth came one of the great mass migrations of Chinese history. Millions of northerners sought safety south of the Yangtze, having taken to the road with the rest of their village or their family, or simply alone. The scale of this mass migration can at present only be estimated. One suggestive clue is the nearly 1.4 million household decline in the census igures of the most affected areas of north and central China between c.1080 and 1162. This drop, even when allowing that many of these would have died in the mayhem of light, conceivably yields a total of 5 million refugees; that is, no less than a seventh of all of north China’s population in c.1080. Since this igure excludes refugees from other areas of the north, the actual igure was most likely far higher. "

So they do estimate the norths population to have been around 35 million before the Jin conquest. They do also extensivly write about something else of interest to this question. The south was suffering a long series of famines and other catastrophes from the 1150s to 1200, which caused a lot of death and tumult. The economy did ajust over time, but it might have caused the population drop and migration to the north we have been speculating about here. Still I think a major reason for the strange numbers is the higher quality of the Jin census vs that of the southern Song, which let the Jin register more people. This would make the S.Song population arround 90 million by 1220 which probably did cause Bannisters large figure (60 million in Jin + Xia and 90 million in Song). MySinicization argument was based on the Jin creating a well working administration in the south, which did take time.
 
Sep 2012
1,121
Taiwan
Do you have a page number for that CHC reference? I'd like to check it out :) It's naturally a contested issue, but I'd say don't write Deng off entirely; he is one of the premier experts in Chinese economic and population history, and well versed in this period in particular. It is easy to say that southward migration is what most often happened, so it is likely it happened here in the Jin too, but as we've seen, that assumption creates further problems: it reduces a Jin population which is already considered to have risen from 1100 figures, and increases a S. Song population which we know declined. It isn't of course to say that northerners affected by the war didn't flee south, only that a net migration north overall must have benefited Jin in the long run.

And yes, there is a massive discrepancy between the Song and Jin census data, the latter being much more comprehensive and authoritative. One of the issues at the heart of these questions is differing interpretations of how to actually use the Song census data, since it can't be directly compared with the Jin figures on account of the fact it only includes men, and massively undersells the amount of people in a household. That's before we even get into whether or not the data is actually accurate in the first place; namely, did it even account for 100% of the male population? How the censuses were actually conducted and their potential shortcomings is well covered in Deng's Unveiling China's True Population Statistics which I cited above; one reason I'm more inclined to trust his figures is because he compensates for this, which few others seem to have done, instead just giving rough estimates based on expected household numbers.